makermondayfandomcom-20200215-history
The Internet of Clothes
Outline How many clothes do you have sitting in your wardrobe that you never wear? A WSJ article suggested that it was as high as 80%. Yet clothes manufacturing is part of a global problem of labour exploitation (40m workers, mostly women) and environmental damage (we consume 4x more clothes than 20 years ago). Owning clothes you never wear is part of the problem. The Internet of Clothes makes us aware of our clothes usage. Each garment will remind, even plead to be worn and if you don't they will automatically offer themselves to charity for re-use. The project objectives is a proof of concept that highlights our wasteful, damaging ownership of clothes and how we can connect these objects to the Internet of Things. For up-to-date project progress please see our Tumblr page: http://makermondaybrum.tumblr.com/ Collaborators Mark Brill (Project Lead) Alexa Hartwell (Partnerships) Adam Woodall (Tech) Giada Totaro (Artistic Consultant) Contact details mark.brill@bcu.ac.uk Objectives * To make clothes 'aware' of how they are used and to offer themselves up to charity if not worn * The project will be developed as a proof of concept - this will be a wardrobe with example clothes that will run on a condensed time-frame (for example, 1 minute=1day) and can be shown in an exhibition space * Each item will be tagged, using RFID. The owner will then categorise each item by naming them, identifying their use (eg rain wear, formal wear) and their likely frequency of wear * As clothes are taken out of the cupboard or drawer an RFID reader will identify that they have been used * A database and algorithm will combine with open weather data to create a set of reminders to the user, for example, a raincoat might send the following message: 'It looks like rain today. Time to wear me? #raincoat'. * Usage reminders will be send via email (or another messaging channel) and Twitter - each item will generate its own hashtag * Clothes not worn for a period based on the user identified frequency will email and Tweet to a charity clothing organisation (for example @OxfamFashion) who will send a mailing envelope for return Development The current state is in first prototype. So far, the concept and architecture has been largely established, we have tested the RFID/laundry tags, completed the technical spec and app wire frames. One challenge is that as clothes don't plug in, so the garments will need to use a battery or passive device. I had considered Beacons (as in iBeacons) but cost, battery life and size. The solution is in the form of RFID. Currently HF RFID (13.56 MHz) seems to be the low-cost option for tags/readers and cost (above all). In practice though the clothing will have to be passed within an inch of the the device in order to be read Option 2 is to use UHF which would work to a more handy 10cm - it's really just cost as a decent UHF reader is ten times the HF price. I see the system working as follows: * Users will set up and account/login using the Twitter account chosen for the clothing items as the username/verification and their email address in user preferences * User management of the data will be via a web-based interface (I'd probably like to create an app as well, but given resources, something HTML5 based will probably do the job) * Each clothing items is chipped using 'laundry' tag - essentially a discreet, washable RFID chip * The RFID reader can be driven by Arduino/Raspberry Pi, making for a small, fairly discreet box either on the wardrobe or bedroom door * Once chipped, each item of clothing will be classified by the owner into a database with the following fields * Chip ID number * Name/hastag - for example, #blackraincoat * Frequency of usage (per month or specific occasional) * Seasonal (summer, spring, winter autumn) * Open data for weather will come via Met Office (for the UK): http://www.metoffice.gov.uk/datapoint or another open data source * The algorithm will be programmed on settings based on frequency of usage/season and current weather * It has been suggested to code this with Python and Pandas database library * For the proof of concept, we will need the support of a charity willing to take email requests from the clothes Equipment * Sparkfun RFID Starter Kit or similar MIFARE reader (the cheap option) * Laundry Tag MIFARE 1K (13.56 MHz) * Raspberry Pi + Power Source * Server space to run the database/web interface * A wardrobe (something not to large) * Some example clothes for the proof of concept Support needs * Developer: my biggest challenge is lack of programming skills to build the database - probably simple for someone with the right skills. I think the interface is best done as HTML5 - I could probably do with a bit of dev help there (but could muddle my way through). * Creative and Build: It would be nice to get some creative support on the overall concept and design for the proof of concept - so how the wardrobe will look, the RFID reader box, the look of the interface and copy writing for the clothes messages Milestones #Test the RFID tags and reader # Set up the system with Raspberri Pi to read the tag numbers # Develop the database/algorithm # Create the user interface, Tweets and emails # Build the wardrobe assembly # Assemble, test and demonstrate the project